[]Trends in consumer goods are leaning towards thinner, more compact products with higher performance than their predecessors. Today’s consumer wants lightweight, portable, battery-operated gear that is powerful enough to satisfy their needs for speed and ease of use. One of the biggest challenges facing product design is thermal management. Case temperature limits for products that come in contact with user’s skin place constraints on internal power dissipation and increase the need for improved thermal management. With the introduction of locally processed AI and higher compact portable device processing loads thermal issues are critical for avoiding throttling. µCooling: A new device has been invented that changes the way thermal engineers are approaching system cooling design. µCooling is a new concept in air movers. It is small, and thin and is a complete departure from rotating mass fans. This MEMS-based invention can move air through very small spaces, used directly on a chip or package, placed remotely through tiny ducts, or mounted outside of a product. In fact, the versatility of µCooling changes the way we can manage heat at the package, SoC, or bare-chip level. This talk will introduce µCooling, a MEMS-based, all-silicon, micro-sized air pump. The device will be described in terms of how it works, performance characteristics, and application examples. Speaker(s): Tom Tarter, Virtual: https://events.vtools.ieee.org/m/500418

[]The rapid rise in power density and complexity of electronic systems has made thermal management a critical challenge for ensuring reliability, performance, and sustainability. Artificial Intelligence (AI) offers transformative opportunities to address this challenge by enabling data-driven modeling, optimization, and predictive control of cooling systems. By integrating AI with experimental and physics-based approaches, adaptive models can be developed to capture transient thermal behaviors, and optimize system-level energy efficiency. This forms the foundation for digital twins, virtual replicas that continuously interact with their physical counterparts to provide system specific real-time monitoring, and data driven decision support. In this talk, I will present recent and ongoing research activities at ES2 Binghamton on AI-enabled thermal management design, with emphasis on cooling solutions for high-power chips in data centers. I will further highlight how these developments serve as a pathway towards creating digital twins, dynamic virtual replicas that integrate real-time data, physics, and AI to enable system-level monitoring, prediction, and optimization. Together, these advancements pave the way for reliable, energy-efficient, and sustainable electronic systems. Speaker(s): Srikanth Rangarajan, Virtual: https://events.vtools.ieee.org/m/504510

[] Requirements on the high-performance compute (HPC) systems from AI workloads necessitates transition to larger package sizes with 2.5D to 3.5D integration and density scaling at every level in the stack. Several competing packaging architectures are emerging to solve the compute and power efficiency challenge presented by AI workloads. Each presents unique lithography challenges such as >100×100 field size, large chip placement deviations, fine lines and tight overlay warped substrates. The conventional lithography tools are incapable of meeting all the requirements to achieve scaling. The talk will preview Applied Materials’ Digital Lithography Technology (DLT) which enables highest resolution at production throughputs while ensuring CD uniformity and overlay accuracy across the entire panel. Speaker(s): Niranjan Khasgiwale, Virtual: https://events.vtools.ieee.org/m/502777